Does earlier diagnosis and treatment of brain tumours matter? Time-to-treatment intervals and tumour size at detection; impact on survival, recurrence, inpatient length of stay and neurological deficit

BackgroundThe importance of timely diagnosis of brain tumours is well recognised but how this translates to impacts on important patient outcomes is not well described. This study aimed to quantify the effect of time-to-treatment interval and tumour size at diagnosis on six patient outcomes: tumour-specific survival, overall survival, new or worsened neurological deficit, 30-day mortality, recurrence and inpatient length of stay in 12 months following diagnosis. MethodsA retrospective cohort study including 1196 patients from Southeast Scotland diagnosed with brain tumours between 2010 and 2020. Regression methods (Cox, Logistic & Linear) used to estimate per day impact of time-to treatment and tumour size impact on outcomes. ResultsThe mean time from first presentation to treatment was 161 days (median: 57 days) with wide variance observed (standard deviation: 330 days). Time-to-treatment results showed an association between longer intervals and reduced mortality. The mean tumour size was 4.06cm (SD: 1.75cm). Each 1cm increase in tumour size increased tumour-specific mortality risk by approximately 11% (HR: 1.11, 95%CI: 1.06-1.16), increased all-cause mortality by 6% (HR: 1.06, 1.02-1.11), increased the expected inpatient stay by 1.5 (0.2-2.8) days, and the odds of new or worsened neurological deficit by 11% (OR: 1.11, 1.01-1.21). ConclusionsLarger tumour size was consistently associated with increased hazard ratios for mortality. Applying these estimates together with estimates of mean tumour growth rates available in the literature, if a 6cm Glioblastoma tumour were diagnosed 1 month earlier then we would expect a 18-28% reduction in hazard of brain tumour mortality. FundingDxcover Ltd. Research in ContextO_ST_ABSEvidence before this studyC_ST_ABSBrain tumour is frequently diagnosed in an emergency setting, with multiple primary care visits often preceding a diagnosis, suggesting opportunities for earlier diagnosis. Previous studies have produced conflicting findings regarding the impact of diagnostic delays on survival and other patient outcomes. Some suggest that longer intervals are associated with worse prognosis, while others report a paradoxical association where delayed diagnosis appears to correlate with improved survival, the waiting time paradox. Prior research has largely focused on the prognostic value of tumour size rather than its role as a mediator of diagnostic delay, limiting causal inference regarding the benefits of earlier detection. Added value of this studyThis study addresses gaps in the literature by examining tumour size as a key mediating factor in the relationship between diagnostic timing and patient outcomes. Using a large, well-curated dataset of 1,196 patients from Southeast Scotland, we demonstrate that tumour size at diagnosis is strongly associated with mortality risk, neurological deficits, inpatient stay, and recurrence. A one-month earlier diagnosis of a 6cm glioblastoma, for example, could reduce tumour size sufficiently to lower the hazard of brain tumour mortality by 18- 28%. By disentangling tumour size effects from time-to-treatment intervals, our study provides a clearer framework for evaluating early detection strategies in neuro-oncology. Implications of all the available evidenceThese findings reinforce the potential impact of earlier diagnosis for brain tumour patients. Building upon the accumulated evidence that tumour size is a critical factor influencing prognosis, this study is the first to robustly estimate the causal association. Earlier detection, through influence on tumour size at diagnosis, may offer measurable survival benefits and potential reductions in healthcare resource use. Given the observational nature of this study, further prospective trials and mathematical modeling are needed to assess the impact of early diagnosis interventions at a population level.